QUB | Archaeology and Palaeoecology | The 14Chrono Centre

Manual for psimpoll and pscomb

Menu

Menu N: confidence intervals

Menu N7: Attempts to model dates without -ve acc. rates
As discussed under menu N6, there can be a problem with excluding negative accumulation rates from modelled age-depth sequences. Each modelled age must be older than the one below it. If, by chance, an age is taken from the old tail of the one of the age distributions, then next down must be older still, and this can get increasingly more difficult, especially where age probability distributions overlap substantially. The default action is to try 1,000,000 times at each age to obtain a kodelled value older than the previous value. If this fails, that particular modelled sequence is abandoned, and the whole sequence re-done. The number of attempts should be as high as possible, limited by the amount of time available, which is dependent on the speed of the computer system being used. The default works well for me, but I recommend trying other values, especially if your sequence of ages has substantial overlaps. The upper allowed value is the maximum allowed value for a `long integer', and may be even higher (but not lower) than the value shown below.

[Allowed range 1-2147483647 (at least)]

Return to menu N.


Back to contents page
Copyright © 1995-2007 K.D. Bennett

Archaeology and Palaeoecology | 42 Fitzwilliam St | Belfast BT9 6AX | Northern Ireland | tel +44 28 90 97 5136


Archaeology and Palaeoecology | The 14Chrono Centre | URL http://www.qub.ac.uk/arcpal/ | WebMaster